The Clinical Utility of lncRNAs and Their Application as Molecular Biomarkers in Breast Cancer

Abstract

Given their tumor-specific and stage-specific gene expression, long non-coding RNAs (lncRNAs) have demonstrated to be potential molecular biomarkers for diagnosis, prognosis, and treatment response. Particularly, the lncRNAs DSCAM-AS1 and GATA3-AS1 serve as examples of this because of their high subtype-specific expression profile in luminal B-like breast cancer. This makes them candidates to use as molecular biomarkers in clinical practice. However, lncRNA studies in breast cancer are limited in sample size and are restricted to the determination of their biological function, which represents an obstacle for its inclusion as molecular biomarkers of clinical utility. Nevertheless, due to their expression specificity among diseases, such as cancer, and their stability in body fluids, lncRNAs are promising molecular biomarkers that could improve the reliability, sensitivity, and specificity of molecular techniques used in clinical diagnosis. The development of lncRNA-based diagnostics and lncRNA-based therapeutics will be useful in routine medical practice to improve patient clinical management and quality of life.

Keywords: biomarkers; breast cancer; lncRNA.

Figure 1

Workflow for lncRNA validation as biomarkers with clinical utility and application. The process of implementing a lncRNA-based biomarker consists of 5 principal steps: discovery, validation, verification, clinical application, and clinical utility [97]. In the discovery step, the objective is to select a lncRNA (or a set) differentially expressed in the condition of interest, like treatment response, and implement it in selected patients. In the validation step, the sample size is increased to determine lncRNA robustness to define the clinical condition of interest and follow the sample size calculation recommended in [20]. In the verification step, lncRNA expression is determined by a clinical laboratory technique, such as qPCR, to verify its viability to be detected in the clinical routine. In the clinical application step, the functionality of the lncRNA as a biomarker for diagnosis, prediction, or prognosis is determined by assessing its sensibility and specificity [20]. Finally, in the clinical utility step, the accuracy of the lncRNA as biomarker is tested in a larger sample size and could be included in clinical trials. (A) PCA3 is an example of a lncRNA that has been validated for clinical application in prostate cancer diagnosis because it represents an FDA-approved lncRNA for clinical purposes. It was discovered from a sample size of 11 patients in the discovery phase [49] and 507 male patients were included in the validation phase in a clinical trial [53]. Additionally, PCA3 is associated with a sensitivity ranging from 54% to 82% and a specificity range of 56.3% to 89%, which justifies its use in clinical practice [50]. Although PCA3 was identified by Northern blot technique [49], it has been validated in other studies by high throughput sequencing technologies, which is the principal tool for current biomarker discovery [98]. (B) EPIC1 is a lncRNA that was identified from the analysis of 6475 tumor samples in the discovery phase and 534 samples in the validation phase [99]. However, it has not yet been verified as a biomarker for clinical utility or for clinical application in the prognosis of breast cancer. (C) GATA3-AS1 is a lncRNA proposed as a potential clinical biomarker in predicting treatment response in breast cancer because it has been demonstrated by Zhang et al. that it is overexpressed in breast neoplasia in a differential expression analysis for 85 paired tumor-normal samples and 830 tumor samples in the discovery phase. For the validation step, 50 paired tumor-normal samples and 23 healthy samples were included [14]. Recently, Contreras-Espinosa et al. also identified this lncRNA by a machine learning approach in a sample size of 11 patients for the discovery step and 68 patients for the validation step, which demonstrated its utility as a predictive biomarker [17]. However, it has not been validated for clinical application yet, as other lncRNA which have been proposed as molecular biomarkers for treatment response prediction in breast cancer, like HOTAIR [8] and MALAT1 [36], have not been included in clinical trials for the analysis of their applicability in diagnosis. lncRNA: long non-coding RNA; ISH: in situ hybridization; PCR: Polymerase Chain Reaction; qPCR: quantitative PCR. Created in BioRender.com.